The Micro-Structural Characterization Of Thermosonic Cu-Al Intermetallic Compounds And Modelling Of Its Interface Stress.
Faculty of Manufacturing Engineering
THE MICRO-STRUCTURAL CHARACTERIZATION OF
THERMOSONIC Cu-Al INTERMETALLIC COMPOUNDS AND
MODELLING OF ITS INTERFACE STRESS
Chua Kok Yau
Doctor of Philosophy
2015
THE MICRO-STRUCTURAL CHARACTERIZATION OF THERMOSONIC CuAl INTERMETALLIC COMPOUNDS AND MODELLING OF ITS INTERFACE
STRESS
CHUA KOK YAU
A thesis submitted
in fulfillment of the requirements for the degree of Doctor of Philosophy
Faculty of Manufacturing Engineering
UNIVERSITI TEKNIKAL MALAYSIA MELAKA
2015
DECLARATION
I declare that this thesis entitled “The Micro-Structural Characterization of Thermosonic
Cu-Al Intermetallic Compounds and Modelling of Its Interface Stress” is the result of my
own research except as cited in the references. The thesis has not been accepted for any
degree and is not concurrently submitted in candidature of any other degree.
Signature
:
...............................................
Name
:
...............................................
Date
:
............................................
APPROVAL
I hereby declare that I have read this thesis and in my opinion this thesis is sufficient in
terms of scope and quality for the award of Doctor of Philosophy.
Signature
:……………. .....................................................
Supervisor Name
:……………………… ......................................
Date
:…………………….. ........................................
DEDICATION
To my beloved parents, wife and children.
ABSTRACT
Thermosonic bonding of the Cu wire on Al bond pad is a common technology used in
semiconductor industry. However, recent research show voids formation at this bonding
interface on micro-chip, after an annealing treatment of High Temperature Storage (HTS).
This voids formation is believed due to the volumetric changes of intermetallic compounds
(IMCs) formed at the bonding interface. In previous research, effects of Cu free-air-ball
and bonding temperature with high temperature storage (HTS) treatment on Cu-Al bonding
interface are unclear. Besides, previous research provides inconclusive knowledge in the
evolution of Cu-Al bonding interface due to inconsistent observations and variations in the
bonding parameters. Research with statistical approach could be useful to address this
limitation, however, it is yet to be established for Thermosonic Cu-Al interconnection.
Besides, the void formation due to volumetric changes of IMC is discussed only
qualitatively. A quantitative stress analysis could close the gap of research. Objectives of
this research are (1) to analyse the correlation of wire bonding parameters, the interfacial
micro-structure change and mechanical strength of the synthesized Cu-Al bonding
interface, (2) to propose a theoretical model that describe quantitatively the stress due to
volumetric changes originated from Cu-Al phase evolution, (3) to evaluate the stress
generated by Cu-Al phase evolution at the bonding interface and its correlation to the void
formation. Micro-structural characterizations were focused on crystallographic,
compositional and mechanical analyses. It was found that bonding temperature resulted in
an exponential increment for initial overall IMC thickness and average Cu content of the
phases formed at the bonding interface. Moreover, HTS increase the overall IMC thickness
by volume diffusion mechanism. The relationship between parameters, mechanical ball
shear strength and IMC thickness were obtained statistically. A mathematical stress model
based on assumptions of isotropic and elastic binary solid-solution was proposed. This
model enabled an estimation of interfacial stresses from compositional measurements. It
was found that the stress developed by interfacial Cu-Al IMC generally increased with the
bonding temperature. Besides, forming gas supply was found to be less significant to affect
the stress development, due to the oxide layers did not hinder much the interdiffusion of
Cu and Al atoms. However, with HTS, the growth of Cu rich IMC increased the stress and
caused gap within copper oxide layer. This work addressed the research gaps and offered a
better understanding of the fundamental of Thermosonic Cu-Al interconnection. The
results of the stress modelling could be a useful failure analysis technique for
implementing Cu wire in the industry.
i
ABSTRAK
Ikatan wayar kuprum (Cu) pada pad aluminium (Al) secara Thermosonic adalah teknologi
biasa digunakan dalam industri semikonduktor. Tetapi, kajian baru-baru ini mendapati
bahawa pembentukan ruang kosong di antaramuka ikatan, selepas penyepuhlindapan
penyimpanan suhu tinggi (PST). Pembentukan ruang kosong ini dipercayai disebabkan
tekanan terjana daripada perubahan isipadu sebatian antara logam (SAL) yang terbentuk di
antaramuka ikatan. Berdasarkan kajian sebelum ini, kesan-kesan ‘bebola-udara-bebas’
wayar kuprum, suhu ikatan dengan PST pada antaramuka ikatan Cu-Al adalah tidak jelas.
Selain itu, kajian sebelum ini tidak memberikan pengetahuan berkesimpulan tentang
revolusi antaramuka ikatang Cu-Al. Ini disebabkan pemerhatian-pemerhatian yang tidak
selaras and parameter-parameter ikatan yang bervariasi. Penyelidikan berdasarkan statistic
adalah berguna untuk menangani batasan ini. Walaubagaimanapun, cara ini belum
ditubuhkan untuk system Cu-Al Thermosonic. Selain itu, pembentukan ruang kosong
disebabkan perubahan isipada SAL telah dibincang secara kualitatif sahaja. Analisis
tekanan secara kuantitatif amat diperlukan. Objektif kajian ini termasuk: (1) untuk
menganalisis korelasi antara pelbagai parameter tersebut, perubahan struktur mikro and
kekuatan mekanik antaramuka Cu-Al disintesis, (2) untuk mencadangkan model teori yang
menghuraikan tekanan disebabkan perubahan isipadu fasa Cu-Al, (3) untuk menilai
tekanan terjana di antaramuka ikatan Cu-Al and hubungannya dengan pembentukan ruang
kosong di situ. Pemcirian struktur dari segi kristalografi, komposisi and mekanik diberi
tumpuan dalam kerja penyelidikan ini. Suhu ikatan didapati membawa kesan kepada
peningkatan tebal SAL awal (secara eksponen) and kandungan Cu dalam SAL ini di
antaramuka ikatan. Lagipun, PST didapati meningkatkan ketebalan keseluruhan SAL
dengan mekanisme peresapan kekisi. Hubungan antara parameter, kekuatan mekanik and
ketebalan SAL diperolehi scara statistik. Model matematik berdasarkan andaian system
logam binari isotropic dan elastik dicadangkan. Model ini membolehkan penganggaran
tekanan antaramuka daripada ukuran komposisi. Tekanan terjana oleh SAL Cu-Al secara
umumnya meningkat dengan suhu ikatan. Bekalan ‘forming gas’ didapati kurang penting
untuk memberi kesan dalam pengembangan tekanan. Ini disebabkan oleh lapisan oksida
berada di antaramuka ikatan tidak menghalang peresapan atom-atom Cu dan Al. Tetapi,
dengan PST, pertumbuhan SAL yang kaya dalam kandungan Cu meningkatkan magnitud
tekanan dan menyebabkan pembentukan jurang dalam lapisan kuprum oxida. Kerja ini
membolehkan pemahaman asas yang lebih baik dalam bidang ikatan Cu-Al Thermosonic.
Keputusan model tekanan berguna sebagai teknik analisis kegagalan untuk pelaksanaan
wayar Cu dalam industry semikonduktor.
ii
ACKNOWLEDGEMENTS
First and foremost, I would like to take this opportunity to express my sincere appreciation
to my main supervisor, Associate Professor Dr. T. Joseph Sahaya Anand, and cosupervisor, Associate Professor Dr. Mohd Warikh Bin Abd Rashid, from the Faculty of
Manufacturing Engineering Universiti Teknikal Malaysia Melaka (UTeM) for their
essential supervision, continuous support and encouragement towards the completion of
this thesis. I would also like to thank my external co-supervisor, Associate Professor Dr.
Azman Jalar, the deputy director of Institute of Microengineering and Nanoelectronics
(IMEN) from Universiti Kebangsaan Malaysia (UKM), for his insightful advice and
continuous support for this work.
Particularly, I would like to thank Ms. Hng May Ting, Ms. Ng Mei Chin, Dr. Lim Boon
Huat, Mr. See Beng Keng, Mr. Ong Meng Tong, Mr. Lee Cher Chia for contributing
materials and processings from industry. Also, supports from Mr. Yong Foo Khong for
scanning transmission electron microscope (STEM) analysis are highly appreciated. This
research could hardly be completed without your supports.
I would also like to express my deepest gratitude to Mr. Azhar Shan Bin Abu Hassan and
Mr. Hairulhisham Bin Rosnan, from material laboratory, Faculty of Manufacturing
Engineering, for their supports and efforts in laboratorial works.
Special thanks to all my parents, wife, and friends for their moral supports along the
journey to complete this research. Last but not least, many thanks to everyone who had
been contributed directly or indirectly to the success of this research.
iii
TABLE OF CONTENTS
Page
DECLARATION
DEDICATION
ABSTRACT
ABSTRAK
ACKNOWLEDGEMENT
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES
LIST OF APPENDICES
LIST OF ABBREVIATIONS
LIST OF SYMBOLS
LIST OF PUBLICATIONS
i
ii
iii
iv
viii
xi
xviii
xix
xxi
xxiii
CHAPTER
1 INTRODUCTION
1.1. Background
1.2. Problem Statement
1.3. Objective
1.4. Scope
1.5. Chapter Outline
2
1
1
4
5
5
6
LITERATURE REVIEWS
2.1. Wire Bonding Technology
2.2. Bond Tool and Geometry of The Bonded Wire
2.3. Wire Bonding Process
2.3.1. Ultrasonic Wire Bonding
2.3.2. Thermosonic Wire Bonding
2.4. Wire Materials
2.5. Bond Pad Materials
2.6. High Temperature Storage (HTS) Treatment
2.7. Thermosonic Cu-Al Bonding Technology and Its Limitations
2.7.1. Cu-Al Intermetallic Compound
2.7.2. Cu-Al Phase Diagram
2.7.3. Bonding Temperature and Its Effects on The Growth of
Interfacial Cu-Al IMC
2.7.4. Forming Gas Supply
2.8. Common Analytical Tools For Assessing Cu-Al Bonding Interface
2.8.1. Field Emission Scanning Electron Microscope
2.8.2. Transmission Electron Microscope
2.8.3. Energy Dispersive X-ray
2.8.4. Nano Beam Diffraction
2.8.5. X-ray Diffraction
iv
8
8
10
13
14
16
17
21
23
24
25
26
31
33
35
36
40
43
46
47
3
4
2.8.6. Mechanical Ball Shear Test
2.9. Stress Analysis
2.10. Statistical Approach
2.10.1. Taguchi Method
2.10.2. Analysis of Variance
2.10.3. Regression analysis
2.11. Summary
50
51
55
55
58
61
63
METHODOLOGY
3.1. The Synthesis of Cu Wire-Al Bond Pad Samples
3.1.1. Design of Experiment
3.2. Sample Preparation
3.2.1. Mechanical Cross-Section for Microscopic Analysis
3.2.2. Focused Ion Beam (FIB) Milling for Lamella Extraction
3.2.3. Sample Collection of Cu Wire-Al Bond Pad for XRD
Measurements
3.3. Characterizations of Cu-Al Samples
3.4. Modelling of Stress Developed at Cu-Al Bonding Interface
3.4.1. Equations for Thermal Misfit and Diffusion-Induced Stresses
3.4.2. Predictions of Material Properties of Solid Solution
3.4.3. Relationship between Concentrations (C s , C i and x)
3.4.4. Partial Molar Volume of Solid Solution
3.5. Summary
66
68
70
74
75
77
78
RESULTS
4.1. Effects of HTS Toward the Microstructures of Cu-Al Bonding Interface
4.1.1. Cu-Al Samples Synthesized At 100°C
4.1.1.1. Imaging Analysis of Cu-Al Bonding Interface by
FESEM
4.1.1.2. Compositional Analysis of Cu-Al Bonding Interface by
EDX
4.1.1.3. STEM, EDX and NBD Analyses on Selected Cu-Al
Samples
4.1.1.4. Total IMC Thickness Measurement Analysis
4.1.1.5. Mechanical Ball Shear Test Analysis
4.1.2. Cu-Al Samples at Bonding Temperature of 280°C
4.1.2.1. FESEM Imaging Studies of Cu-Al Bonding Interface
4.1.2.2. EDX Compositional Studies of Cu-Al Bonding
Interface
4.1.2.3. STEM and EDX Analyses on Samples Before and
After 1000 hours of HTS Treatment
4.1.2.4. Total Cu-Al IMC Thickness Analysis
4.1.2.5. Mechanical Ball Shear Test Analysis
4.1.3. Cu-Al Samples Synthesized at Bonding Temperature of 400°C
94
94
95
95
v
80
82
83
87
89
90
93
101
106
113
114
115
115
122
127
137
138
139
4.2.
4.3.
4.4.
4.5.
4.6.
4.7.
4.8.
5
4.1.3.1. Morphological Studies of Cu-Al Bonding Interface by
FESEM
4.1.3.2. Compositional Studies of Cu-Al Bonding Interface by
EDX
4.1.3.3. STEM and EDX Analyses on Selected Cu-Al Samples
Synthesized at 400°C
4.1.3.4. Total IMC Thickness Measurement Analysis
4.1.3.5. Mechanical Ball Shear Test Analysis
Effect of Forming Gas Supply Towards The Micro-Structure of
Interfacial IMC
4.2.1. Cross-Sectional FESEM Imaging Analysis at Cu-Al Bonding
Interface
4.2.2. STEM and EDX Analyses at Cu-Al Bonding Interface
4.2.3. Mechanical Studies of Ball Shear Test
Mechanical Ball Shear Performance Analysis by Taguchi approach
STEM and EDX Analyses at Cu-Al Bonding Interface for Taguchi
Sample Matrix
4.4.1. Characterizations on Annealed Sample Synthesized at 100°C
Without Forming Gas Supply
4.4.2. Characterizations on Annealed Sample Synthesized at 280°C
With Forming Gas OFF
4.4.3. Characterizations on Annealed Sample Synthesized at 400°C
Without Forming Gas Supply
XRD Characterization With Two Different Sample Preparation
Techniques
4.5.1. By Mechanical Collection Technique
4.5.2. By Chemical Etching Technique
Theoretical Modelling
4.6.1. Prediction of Elastic Properties of Solid Solution
4.6.2. Prediction of Concentration of Solute and Partial Molar Volume
of Solid Solution
4.6.3. Prediction of Stresses
4.6.4. Stress Analysis to Evaluate the Effect of Bonding Temperatures
4.6.5. Stress Analysis To Assess the Effect of HTS Treatment
4.6.5.1. For Cu-Al Samples Synthesized at 100°C
4.6.5.2. For Cu-Al Samples Synthesized at 280°C
4.6.6. Stress Analysis to Evaluate the Effect of Forming Gas Supply
4.6.7. Stress Analysis Based on Taguchi Orthogonal Matrix
Applications of Taguchi Method on The Stresses Generated at Cu-Al
Bonding Interface
Summary
DISCUSSION
5.1. IMC Growth at The Bonding Interface and Its Correlation to The
vi
139
145
149
157
159
160
160
162
174
175
179
179
184
187
191
191
199
201
201
205
207
208
211
211
212
214
216
220
222
225
225
Mechanical Bond Strength
Relationship Between Ball Shear Strength and Bonding Parameters
Chemical Composition Distribution Across The Bonding Interface
Grain Boundary Diffusion at Low Bonding Temperature
The Presence of Oxide Layers at Cu-Al Bonding Interface
Formation of Gap Within The Oxide Layer at Cu-Al Bonding Interface
Possible Mechanism of The Formation of Uncommon Features at Cu-Al
Bonding Interface
5.8. Summary
5.2.
5.3.
5.4.
5.5.
5.6.
5.7.
6
CONCLUSION AND RECOMMENDATIONS
STUDIES
6.1. Conclusion
6.2. Contributions To Knowledge
6.3. Recommendations For Future Studies
FOR
229
230
232
233
237
240
245
FUTURE 247
247
248
249
251
264
REFERENCES
APPENDICES
vii
LIST OF TABLE
Table
Title
Page
2.1
Comparison between wire bonding technologies (Harman, 2010; Pan and
Fraud, 2004; Charles, 2007; Lum et al., 2006; Xu et al., 2008)
11
2.2
Material properties comparison between Cu and Au (Calister, 2004).
20
2.3
Micro-structural and physical properties of Cu-Al phases (Xu, 2010).
28
2.4
Wire bonding parameters applied in some of the reports focus on Thermosonic
Cu-Al system.
32
2.5
d hkl for various crystal systems and their relationship with Miller indices.
49
2.6
Orthogonal array (L9) (Puh et al., 2012)
56
2.7
S/N equations for different quality characteristic (Nalbant et al., 2007)
57
2.8
Standard two way AVONA arrangement (Sahai and Ageel, 2000).
59
2.9
Examples of critical F values based on α=0.10 (Walck, 2007).
61
3.1
Optimized Thermosonic wire bonding parameters used to synthesize Cu-Al
samples.
69
3.2
Full factorial sample matrix of Cu-Al samples.
71
3.3
Taguchi L9 Orthogonal Matrix used to synthesize Cu-Al samples.
72
3.4
Sample matrix of Cu-Al samples for direct comparisons.
73
3.5
Sample Matrix to evaluate the effect of HTS treatment on Cu-Al samples.
Forming gas supplies were turned ON.
74
viii
4.1
Summary of EDX measurements of sample synthesized at 100°C.
102
4.2
Summary of predicted Cu-Al phases from FESEM EDX for sample synthesized
at 100°C.
104
4.3
Summary of compositional analysis on as-synthesized sample produced with
100°C.
109
4.4
Summary of compositional analyses on sample synthesized at 100°C after 1000
hours of HTS.
112
4.5
Summary of EDX measurements of sample synthesized at 280°C.
123
4.6
Summary of predicted phases for sample synthesized at 280°C.
124
4.7
Summary of compositional analysis of Cu-Al bonding interface interpreted
from Figure 4.19.
130
4.8
Summary of compositional analysis of sample synthesized at 280°C after 1000
hours of HTS.
136
4.9
Summary of EDX measurements of sample synthesized at 400°C.
146
4.1
Summary of predicted phases for sample synthesized at 400°C.
147
4.11
Summary of compositional analysis of Cu-Al bonding interface of the sample
synthesized at 400° before HTS treatment.
152
4.12
Summary of compositional analysis of Cu-Al bonding interface of the sample
synthesized at 400° after 500 hours HTS treatment.
157
4.13
Summary of compositional analysis of the Cu-Al sample synthesized at 280°C
and without forming gas supply.
173
4.14
Results of ball shear measurements of the Cu-Al samples and their S/N ratio.
176
4.15
Response Table for S/N ratio.
176
4.16
Comparison between the predicted and measured S/N ratio and shear strength
of the Cu-Al samples.
177
4.17
ANOVA of ball shear strengths of the Cu-Al samples.
178
4.18
Summary of the compositional analysis on sample S4
183
4.19
Summary of the compositional studies on sample S5.
187
4.20
Summary of the compositional analysis of sample S9.
190
4.21
Summary of crystallographic detail of phase detected.
197
ix
4.22
Material related constants used in the analysis (Calister, 2004).
202
4.23
Results of the predicted stress and their S/N ratio
220
4.24
Response table for the S/N ratio of the stresses predicted.
221
5.1
Features and stresses of various locations at Cu-Al bonding interface of S1.
239
x
LIST OF FIGURE
Figure
Title
Page
2.1
Illustrations of (a) ceramic capillary, (b) cross-sectional view of the tapered tip
of (a). Modified from (CoorsTek, 2014)
12
2.2
Illustrations of (a) ball bond and (b) wedge bond. Modified from (Charles,
2009)
13
2.3
The illustrations for the process flow of Ultrasonic wedge bonding. Modified
from (Harman, 2010).
15
2.4
The illustration for the process flow of Thermosonic ball bonding. Modified
from (Harman, 2010).
18
2.5
Four tiered ball bond management in a micro-chip for graphic processing
applications (Singh et al., 2005).
20
2.6
The appearance of IMC layer in sample synthesized at 280°C after 500 hours
of HTS 175°C.
26
2.7
Binary Cu-Al thermal equilibrium phase diagram (ASM, 1992).
28
2.8
Comparison of FESEM micrograph of Cu FABs synthesized with (a) forming
gas ON, (b) forming gas OFF (Lee et al., 2010).
34
2.9
The design of a common copper kit. (Breach, 2010).
34
2.10
The construction of a conventional SEM equipment (Zhou et al., 2007).
37
2.11
SEM micrographs show the structures of (a) thermionic and (b) field effect
electron sources (Spectral solutions, 2014).
38
2.12
Illustration of signal generations resulted from interactions between incident
electron and the surface of a sample (Fultz and Howe, 2013).
40
2.13
Construction of TEM equipment (Fultz and Howe, 2013).
41
2.14
Schematic of an annular detector in a STEM. Modified from (Fultz and Howe,
2013).
43
2.15
The broadening of energetic electron beam in (a) bulk sample, and (b) in thin
lamella. Adapted from (Fultz and Howe, 2013).
45
xi
2.16
Illustration of EDX measurements taken at Cu, IMC and Al. Overlapping of
the interaction volume and Cu bulk results occurs when EDX measurement is
taken close to the Cu-IMC interface.
An indexed XRD diffractogram of Al thin film (Birkholz, 2006). Inset shows
the similar diffractogram with y-axiz in square-root of count rate.
45
2.18
Mechanical ball shear test configuration (Harman, 2010).
51
3.1
Flow chart of the processes of the experiment.
67
3.2
The illustration of the structure of the synthesized Thermosonic Cu-Al sample.
69
3.3
OM images of cross-sectioned samples with (a) a smooth polished bonding
interface, and (b) a damaged bonding interface.
77
3.4
The geometrical considerations of the multi-layer structure.
85
3.5
The illustration of the process flow of the stress calculation. (a) The
composition measurements, (b) the calculated x Cu profile from (a), (c) the
converted C profile, (d) Predicted elastic material properties, (e) predicted
stresses profile.
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 100°C before HTS treatment.
92
4.2
FESEM micrographs of cross-sectional Cu-Al bonding interface synthesized at
100°C after 200 hours of HTS treatment.
97
4.3
FESEM micrographs of cross-sectional Cu-Al bonding interface synthesized at
100°C after 500 hours of HTS treatment.
98
4.4
FESEM micrographs of cross-sectional Cu-Al bonding interface synthesized at
100°C after 700 hours of HTS treatment.
99
4.5
FESEM micrographs of cross-sectional Cu-Al bonding interface synthesized at
100°C after 1000 hours of HTS treatment.
100
4.6
(a) STEM overview of the lamella of sample synthesized at 100°C before HTS
treatment. The magnified micrographs at selected regions are shown in inset
(b) to (d).
(a) to (d) line scan EDX profiles (i) to (iv) as defined in Figure 4.6,
respectively.
108
4.8
NBD pattern taken from the IMC grain.
109
4.9
(a) STEM overview of the lamella of sample synthesized at 100°Cafter 1000
hours of HTS treatment. (b) to (d) show the magnified micrograph at selected
regions.
(a) to (c) were results of line scan EDX (i) to (iii) as defined in Figure 4.9,
respectively.
111
(a) The plot of total IMC thickness versus HTS durations for sample
synthesized at 100°C, (b) Results of linear regression analysis.
114
2.17
4.1
4.7
4.10
4.11
xii
48
96
108
111
4.12
Box plot of ball shear strength versus HTS durations for samples synthesized
at the bonding temperature of 100°C.
115
4.13
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C before HTS treatment.
117
4.14
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C after 200 hours of HTS treatment.
118
4.15
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C after 500 hours of HTS treatment.
119
4.16
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C after 700 hours of HTS treatment.
120
4.17
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C after 1000 hours of HTS treatment.
121
4.18
(a) STEM overview of the lamella of the Cu-Al bonding interface synthesized
at 280°C before HTS treatment. (b) to (c) show the magnified micrographs at
selected regions.
(a) to (c) were line scan EDX Cu-Al profiles of measurements (i) to (iii) as
defined in Figure 4.18, respectively.
128
(a) STEM overview of the lamella of sample synthesized at 280°C after 1000
hours of HTS treatment. (b) to (d) show the magnified micrograph at selected
regions.
Magnified STEM micrograph revealed the existence of the particles and
stripes features in the IMC with darker. Synthesized conditions: bonding
temperature 280C and after 1000 hours of HTS treatment.
(a) to (f) were line scan EDX profiles (i) to (vi) as defined in Figure 4.20,
respectively.
131
4.23
NBD diffraction pattern that matched the diffration from zone axis [-1 -2 0] of
θ phase.
133
4.24
(a) The plot of total IMC thickness versus HTS durations for the Cu-Al
samples synthesized at 280°C, (b) Results of the corresponding linear
regression analysis.
Box plot of ball shear strength versus HTS durations for the Cu-Al samples
synthesized at the bonding temperature of 280°C.
137
4.26
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 400°C before HTS treatment.
140
4.27
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 400°C after HTS 200 hours.
141
4.28
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 400°C after HTS 500 hours.
142
4.29
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 400°C after HTS 700 hours.
143
4.30
FESEM micrographs of the cross-sectional bonding interface synthesized at
400°C after HTS 1000 hours.
144
4.19
4.20
4.21
4.22
4.25
xiii
128
132
133
138
4.31
(a) STEM overview of the lamella of sample synthesized at 400°C before HTS
treatment. (b) to (d) show the magnified micrograph at selected regions.
151
4.32
(a) to (c) show line scan EDX profiles of measurements (i) to (iii) as defined in
Figure 4.31, respectively.
151
4.33
(a) STEM overview of the lamella of sample synthesized at 400°C after 500
hours of HTS treatment. (b) to (d) show the magnified micrograph of the
selected regions around the bonding interface.
Results of EDX line scan measurements as defined by arrow (i) to (v) in
Figure 4.33.
153
(a) The plot of total IMC thickness versus HTS durations for Cu-Al sample
synthesized at 400°C, (b) the results of the corresponding linear regression
analysis.
Box plot of ball shear strength versus HTS durations for Cu-Al samples
synthesized at the bonding temperature of 400°C.
158
4.37
FESEM micrographs of the cross-sectioned of the as-synthesized Cu-Al
sample produced at 280°C and without the forming gas supply.
161
4.38
Bright field TEM micrograph of the Cu-Al sample synthesized at 280°C and
without forming gas supply. (a) to (d) are magnified micrographs that revealed
voids and IMC features. Inset (e) shows a magnified STEM micrograph that
better visualize the present of IMC.
OM images of (a) a symmetrical non-oxidized ball bond and (b) an
asymmetrical and off-centred oxidized ball bond.
163
Bright field TEM micrographs of Cu-Al bonding interface synthesized at
280°C and with forming gas supply ON. (a) to (d) are magnified micrographs
that revealed voids and IMC features at the bonding interface. (c) is a dark
field image.
IMC thickness comparison between sample with and without forming gas
supply.
167
4.42
Line scan EDX results across an IMC layer of (a) sample with forming gas
ON, (b) sample with forming gas OFF.
169
4.43
STEM micrographs of the as-synthesized sample bonded with bonding
temperature 280°C and forming gas OFF. (a) represents the overview
micrograph of the lamella. (b) to (d) are magnified micrographs at selected
region of the lamella.
Composition profiles across the Cu-Al bonding interface as defined by arrows
(i) to (vi) in Figure 4.43.
171
(a) and (b) are the FESEM micrographs for the Cu-Al bonding interface
synthesized with forming gas supply OFF and ON, respectively. Insets show
magnified micrograph in high stress regions for IMC thickness measurements.
The response graph of S/N ratio of ball shear strength with different factors
and levels.
174
STEM micrographs of the sample S4 synthesized with bonding temperature
100°C, forming gas OFF and after 500 hours of HTS treatment. (a) shows the
overview micrograph of the lamella. (b) to (d) are magnified micrographs at
181
4.34
4.35
4.36
4.39
4.40
4.41
4.44
4.45
4.46
4.47
xiv
153
159
165
168
171
177
some selected regions around the bonding interface.
4.48
(a) to (d) show the line scan EDX results measured along arrow (i) to (iv)
defined in Figure 4.47.
181
4.49
NBD diffraction patterns obtained from the IMC. The indexing of the
diffraction patterns matched (a) [1 -1 0] from θ, and (b) [-9 -7 6] from η 2
phases.
STEM micrographs of the sample S5 synthesized at 280°C with forming gas
OFF and after 500 hours of HTS treatment. (a) shows the overview micrograph
of the lamella. (b) and (c) are magnified micrographs at selected regions at the
bonding interface.
(a) to (c) show the line scan EDX results measured along arrow (i) to (iv) as
defined in Figure 4.50.
182
STEM images of sample S5 that was synthesized at 400°C without forming
gas supply and after HTS 1000hours. (a) was the overview micrograph, (b) to
(d) were the magnified micrographs at the selected RoIs.
Compositional studies on sample S9. (a) to (d) were the measurements’ results
taken along arrow (i) to (iv) as defined in Figure 4.52.
188
4.54
(a) The optical image of the wire bonded specimen in leadframe form. (b) The
schematic demonstrating the material configurations of a unit of specimen.
192
4.55
OM image of crushed specimen prepared by the manual collection of the wires
and micro-chips.
193
4.56
(a) FESEM of the crushed Cu-Al specimen, (b) corresponding EDX results
193
4.57
194
4.58
(a) Optical micrograph of a bundle of the specimen consisted of Cu wires and
Al bond pads. (b) magnified optical micrograph showing the existence of Al
under Cu ball bond.
EDX analysis on the bottom side of the ball bond. Al was detected.
4.59
XRD diffractograms for the specimens prepared with three different methods.
196
4.60
(a) The diffractogram from XRD measurement using Noiseless sample holder.
(b) A duplicate of diffractogram from Figure 4.59(b).
198
4.61
SEM EDX validation on the elimination of Cu-Al IMC after etching with
KOH 50%.
200
4.62
Diffractograms after KOH etching with different concentrations.
201
4.63
(a) The reference model of Cu-Al metal couple with a layer of solid solution
sandwiched between the pure metals. (b) The plot of Cu concentration versus
the thickness of the solid solution.
The comparison of reported and predicted values of (a) shear, (b) bulk and (c)
Young’s moduli.
202
4.50
4.51
4.52
4.53
4.64
xv
186
186
188
195
204
4.65
The plot of the predicted solute concentration, C versus fractional
concentration of Cu, x Cu .
206
4.66
The plot of the calculated partial molar volume of the solid solution, Ṽ versus
x Cu .
206
4.67
The calculated distribution of three components of the diffusion induced stress.
207
4.68
(a) STEM images on the IMC formation at the bonding interface. (b) The
corresponding x Cu profile. (c) The corresponding predicted stresses. (i) to (iii)
represent the results of the samples synthesized at 100, 280 and 400°C.
(a) STEM image of S7 synthesized at 100°C after 1000 hours of HTS
treatment. (b) x Cu profile obtained from the EDX measurement defined in (a).
(c) The predicted stress profile along the EDX line across the bonding
interface.
(a) STEM image of the bonding interface of S8 synthesized at 280°C after
1000 hours of HTS treatment. (b) x Cu profile calculated from the EDX
measurement defined in (a). (c) The predicted stress profile from x Cu profile.
(a) STEM image of the bonding interface of S1 synthesized at 280°C before
HTS treatment. (b) x Cu profile calculated from the line scan EDX measurement
defined in (a). (c) The predicted stress profile from x Cu profile.
(a) STEM micrograph of sample S4 synthesized at 100°C without forming gas
supply and after 500 hours of HTS treatment. (b) The calculated x Cu profile
from EDX measurement defined in (a). (c) The predicted stress distribution.
(a) STEM micrograph of sample S5 synthesized at 280°C without forming gas
supply and after 500 hours of HTS treatment. (b) The x Cu profile from the
EDX measurement defined in (a). (c) The calculated stress distribution.
STEM micrograph of sample S6 synthesized at 400°C with forming gas ON
and after 500 hours of HTS treatment. (b) the x Cu profile from the EDX
measurement specified in (a). (c) The predicted stress distribution.
STEM micrograph of sample S9 synthesized at 400°C with forming gas OFF
and after 1000 hours of HTS treatment. (b) the x Cu profile from the EDX
measurement specified in (a). (c) The predicted stress distribution.
IMC thickness versus HTS durations for the bonding temperatures of (a) 100,
(b) 280 and (c) 400°C.
209
5.2
Histogram of the ball shear strength distribution of samples prepared with
different temperatures and HTS durations.
228
5.3
The plot of ball shear strength versus total IMC thickness.
228
5.4
The chemical composition distribution at both central and peripheral RoIs.
232
5.5
Chemical compositional distribution at Cu-Al bonding interfaces of (a) S10,
(b) S8 and (c) S6.
234
5.6
(a) to (d) show the composition profiles of the bonding interface of samples
S1, S4, S5 and S9, respectively.
236
4.69
4.70
4.71
4.72
4.73
4.74
4.75
5.1
xvi
212
213
215
216
218
218
219
226
5.7
The compiled composition profiles and their respective stress profiles of
sample S1.
239
5.8
STEM micrograph of sample synthesized at 280°C with forming gas supply.
(a) and (b) were that before and after 1000 hours of HTS treatment,
respectively. Insets in (b) show the magnified micrographs that reveals the
presence of rod and stripe-like features.
STEM micrograph shows the appearance of stripes and rods-like feature
present at Cu-Al bonding interface.
241
Schematics of proposed mechanism of formations of uncommon features. (a)
the initial thin IMC formation, (b) with prolonged HTS, volume and lateral
diffusions of Cu atoms occurred simultaneously, (c) Upon consumption of the
bond pad, the diffused Cu atoms partially penetrated into Si together with
inversed diffusion to thicken a new IMC layer with higher Cu content, (d) the
formations of stripes, rods-like features and periodic Cu composition were
observed.
STEM micrograph of sample S6. The structure of the IMC formed was
apparently rods-like.
244
5.9
5.10
5.11
xvii
243
244
LIST OF APPENDICES
APPENDIX
A
B
C
TITLE
Derivations of ��0 , �� , ��0 And ��
Derivations of Conversion Equations
Derivation of The Number of Mole of Solute Atoms per
unit Volume, C
xviii
PAGE
264
269
272
LIST OF ABBREVIATIONS
Al
-
Aluminum
ASM
-
American Society for Metals
Au
-
Gold
BGA
-
Ball Grid Array
BST
Ball Shear Test
CTE
-
Coefficient of Thermal Expansion
Cu
-
Copper
EDX
-
Energy Dispersive X-Ray
EFO
-
Electro-Flame-Off
FAB
-
Free-Air-Ball
FESEM
-
Field Emission Scanning Electron Microscope
GB
-
Grain Boundary
HTS
-
High Temperature Storage
IMC
-
Intermetallic Compound
JEDEC
-
Joint Electronic Device Engineering Council
Ni
-
Nickel
SEM
Scanning Electron Microscope
Si
-
Silicon
STEM
-
Scanning Transmission Electron Microscope
W
-
Tungstun
XRD
-
X-Ray Diffraction
xix
THE MICRO-STRUCTURAL CHARACTERIZATION OF
THERMOSONIC Cu-Al INTERMETALLIC COMPOUNDS AND
MODELLING OF ITS INTERFACE STRESS
Chua Kok Yau
Doctor of Philosophy
2015
THE MICRO-STRUCTURAL CHARACTERIZATION OF THERMOSONIC CuAl INTERMETALLIC COMPOUNDS AND MODELLING OF ITS INTERFACE
STRESS
CHUA KOK YAU
A thesis submitted
in fulfillment of the requirements for the degree of Doctor of Philosophy
Faculty of Manufacturing Engineering
UNIVERSITI TEKNIKAL MALAYSIA MELAKA
2015
DECLARATION
I declare that this thesis entitled “The Micro-Structural Characterization of Thermosonic
Cu-Al Intermetallic Compounds and Modelling of Its Interface Stress” is the result of my
own research except as cited in the references. The thesis has not been accepted for any
degree and is not concurrently submitted in candidature of any other degree.
Signature
:
...............................................
Name
:
...............................................
Date
:
............................................
APPROVAL
I hereby declare that I have read this thesis and in my opinion this thesis is sufficient in
terms of scope and quality for the award of Doctor of Philosophy.
Signature
:……………. .....................................................
Supervisor Name
:……………………… ......................................
Date
:…………………….. ........................................
DEDICATION
To my beloved parents, wife and children.
ABSTRACT
Thermosonic bonding of the Cu wire on Al bond pad is a common technology used in
semiconductor industry. However, recent research show voids formation at this bonding
interface on micro-chip, after an annealing treatment of High Temperature Storage (HTS).
This voids formation is believed due to the volumetric changes of intermetallic compounds
(IMCs) formed at the bonding interface. In previous research, effects of Cu free-air-ball
and bonding temperature with high temperature storage (HTS) treatment on Cu-Al bonding
interface are unclear. Besides, previous research provides inconclusive knowledge in the
evolution of Cu-Al bonding interface due to inconsistent observations and variations in the
bonding parameters. Research with statistical approach could be useful to address this
limitation, however, it is yet to be established for Thermosonic Cu-Al interconnection.
Besides, the void formation due to volumetric changes of IMC is discussed only
qualitatively. A quantitative stress analysis could close the gap of research. Objectives of
this research are (1) to analyse the correlation of wire bonding parameters, the interfacial
micro-structure change and mechanical strength of the synthesized Cu-Al bonding
interface, (2) to propose a theoretical model that describe quantitatively the stress due to
volumetric changes originated from Cu-Al phase evolution, (3) to evaluate the stress
generated by Cu-Al phase evolution at the bonding interface and its correlation to the void
formation. Micro-structural characterizations were focused on crystallographic,
compositional and mechanical analyses. It was found that bonding temperature resulted in
an exponential increment for initial overall IMC thickness and average Cu content of the
phases formed at the bonding interface. Moreover, HTS increase the overall IMC thickness
by volume diffusion mechanism. The relationship between parameters, mechanical ball
shear strength and IMC thickness were obtained statistically. A mathematical stress model
based on assumptions of isotropic and elastic binary solid-solution was proposed. This
model enabled an estimation of interfacial stresses from compositional measurements. It
was found that the stress developed by interfacial Cu-Al IMC generally increased with the
bonding temperature. Besides, forming gas supply was found to be less significant to affect
the stress development, due to the oxide layers did not hinder much the interdiffusion of
Cu and Al atoms. However, with HTS, the growth of Cu rich IMC increased the stress and
caused gap within copper oxide layer. This work addressed the research gaps and offered a
better understanding of the fundamental of Thermosonic Cu-Al interconnection. The
results of the stress modelling could be a useful failure analysis technique for
implementing Cu wire in the industry.
i
ABSTRAK
Ikatan wayar kuprum (Cu) pada pad aluminium (Al) secara Thermosonic adalah teknologi
biasa digunakan dalam industri semikonduktor. Tetapi, kajian baru-baru ini mendapati
bahawa pembentukan ruang kosong di antaramuka ikatan, selepas penyepuhlindapan
penyimpanan suhu tinggi (PST). Pembentukan ruang kosong ini dipercayai disebabkan
tekanan terjana daripada perubahan isipadu sebatian antara logam (SAL) yang terbentuk di
antaramuka ikatan. Berdasarkan kajian sebelum ini, kesan-kesan ‘bebola-udara-bebas’
wayar kuprum, suhu ikatan dengan PST pada antaramuka ikatan Cu-Al adalah tidak jelas.
Selain itu, kajian sebelum ini tidak memberikan pengetahuan berkesimpulan tentang
revolusi antaramuka ikatang Cu-Al. Ini disebabkan pemerhatian-pemerhatian yang tidak
selaras and parameter-parameter ikatan yang bervariasi. Penyelidikan berdasarkan statistic
adalah berguna untuk menangani batasan ini. Walaubagaimanapun, cara ini belum
ditubuhkan untuk system Cu-Al Thermosonic. Selain itu, pembentukan ruang kosong
disebabkan perubahan isipada SAL telah dibincang secara kualitatif sahaja. Analisis
tekanan secara kuantitatif amat diperlukan. Objektif kajian ini termasuk: (1) untuk
menganalisis korelasi antara pelbagai parameter tersebut, perubahan struktur mikro and
kekuatan mekanik antaramuka Cu-Al disintesis, (2) untuk mencadangkan model teori yang
menghuraikan tekanan disebabkan perubahan isipadu fasa Cu-Al, (3) untuk menilai
tekanan terjana di antaramuka ikatan Cu-Al and hubungannya dengan pembentukan ruang
kosong di situ. Pemcirian struktur dari segi kristalografi, komposisi and mekanik diberi
tumpuan dalam kerja penyelidikan ini. Suhu ikatan didapati membawa kesan kepada
peningkatan tebal SAL awal (secara eksponen) and kandungan Cu dalam SAL ini di
antaramuka ikatan. Lagipun, PST didapati meningkatkan ketebalan keseluruhan SAL
dengan mekanisme peresapan kekisi. Hubungan antara parameter, kekuatan mekanik and
ketebalan SAL diperolehi scara statistik. Model matematik berdasarkan andaian system
logam binari isotropic dan elastik dicadangkan. Model ini membolehkan penganggaran
tekanan antaramuka daripada ukuran komposisi. Tekanan terjana oleh SAL Cu-Al secara
umumnya meningkat dengan suhu ikatan. Bekalan ‘forming gas’ didapati kurang penting
untuk memberi kesan dalam pengembangan tekanan. Ini disebabkan oleh lapisan oksida
berada di antaramuka ikatan tidak menghalang peresapan atom-atom Cu dan Al. Tetapi,
dengan PST, pertumbuhan SAL yang kaya dalam kandungan Cu meningkatkan magnitud
tekanan dan menyebabkan pembentukan jurang dalam lapisan kuprum oxida. Kerja ini
membolehkan pemahaman asas yang lebih baik dalam bidang ikatan Cu-Al Thermosonic.
Keputusan model tekanan berguna sebagai teknik analisis kegagalan untuk pelaksanaan
wayar Cu dalam industry semikonduktor.
ii
ACKNOWLEDGEMENTS
First and foremost, I would like to take this opportunity to express my sincere appreciation
to my main supervisor, Associate Professor Dr. T. Joseph Sahaya Anand, and cosupervisor, Associate Professor Dr. Mohd Warikh Bin Abd Rashid, from the Faculty of
Manufacturing Engineering Universiti Teknikal Malaysia Melaka (UTeM) for their
essential supervision, continuous support and encouragement towards the completion of
this thesis. I would also like to thank my external co-supervisor, Associate Professor Dr.
Azman Jalar, the deputy director of Institute of Microengineering and Nanoelectronics
(IMEN) from Universiti Kebangsaan Malaysia (UKM), for his insightful advice and
continuous support for this work.
Particularly, I would like to thank Ms. Hng May Ting, Ms. Ng Mei Chin, Dr. Lim Boon
Huat, Mr. See Beng Keng, Mr. Ong Meng Tong, Mr. Lee Cher Chia for contributing
materials and processings from industry. Also, supports from Mr. Yong Foo Khong for
scanning transmission electron microscope (STEM) analysis are highly appreciated. This
research could hardly be completed without your supports.
I would also like to express my deepest gratitude to Mr. Azhar Shan Bin Abu Hassan and
Mr. Hairulhisham Bin Rosnan, from material laboratory, Faculty of Manufacturing
Engineering, for their supports and efforts in laboratorial works.
Special thanks to all my parents, wife, and friends for their moral supports along the
journey to complete this research. Last but not least, many thanks to everyone who had
been contributed directly or indirectly to the success of this research.
iii
TABLE OF CONTENTS
Page
DECLARATION
DEDICATION
ABSTRACT
ABSTRAK
ACKNOWLEDGEMENT
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES
LIST OF APPENDICES
LIST OF ABBREVIATIONS
LIST OF SYMBOLS
LIST OF PUBLICATIONS
i
ii
iii
iv
viii
xi
xviii
xix
xxi
xxiii
CHAPTER
1 INTRODUCTION
1.1. Background
1.2. Problem Statement
1.3. Objective
1.4. Scope
1.5. Chapter Outline
2
1
1
4
5
5
6
LITERATURE REVIEWS
2.1. Wire Bonding Technology
2.2. Bond Tool and Geometry of The Bonded Wire
2.3. Wire Bonding Process
2.3.1. Ultrasonic Wire Bonding
2.3.2. Thermosonic Wire Bonding
2.4. Wire Materials
2.5. Bond Pad Materials
2.6. High Temperature Storage (HTS) Treatment
2.7. Thermosonic Cu-Al Bonding Technology and Its Limitations
2.7.1. Cu-Al Intermetallic Compound
2.7.2. Cu-Al Phase Diagram
2.7.3. Bonding Temperature and Its Effects on The Growth of
Interfacial Cu-Al IMC
2.7.4. Forming Gas Supply
2.8. Common Analytical Tools For Assessing Cu-Al Bonding Interface
2.8.1. Field Emission Scanning Electron Microscope
2.8.2. Transmission Electron Microscope
2.8.3. Energy Dispersive X-ray
2.8.4. Nano Beam Diffraction
2.8.5. X-ray Diffraction
iv
8
8
10
13
14
16
17
21
23
24
25
26
31
33
35
36
40
43
46
47
3
4
2.8.6. Mechanical Ball Shear Test
2.9. Stress Analysis
2.10. Statistical Approach
2.10.1. Taguchi Method
2.10.2. Analysis of Variance
2.10.3. Regression analysis
2.11. Summary
50
51
55
55
58
61
63
METHODOLOGY
3.1. The Synthesis of Cu Wire-Al Bond Pad Samples
3.1.1. Design of Experiment
3.2. Sample Preparation
3.2.1. Mechanical Cross-Section for Microscopic Analysis
3.2.2. Focused Ion Beam (FIB) Milling for Lamella Extraction
3.2.3. Sample Collection of Cu Wire-Al Bond Pad for XRD
Measurements
3.3. Characterizations of Cu-Al Samples
3.4. Modelling of Stress Developed at Cu-Al Bonding Interface
3.4.1. Equations for Thermal Misfit and Diffusion-Induced Stresses
3.4.2. Predictions of Material Properties of Solid Solution
3.4.3. Relationship between Concentrations (C s , C i and x)
3.4.4. Partial Molar Volume of Solid Solution
3.5. Summary
66
68
70
74
75
77
78
RESULTS
4.1. Effects of HTS Toward the Microstructures of Cu-Al Bonding Interface
4.1.1. Cu-Al Samples Synthesized At 100°C
4.1.1.1. Imaging Analysis of Cu-Al Bonding Interface by
FESEM
4.1.1.2. Compositional Analysis of Cu-Al Bonding Interface by
EDX
4.1.1.3. STEM, EDX and NBD Analyses on Selected Cu-Al
Samples
4.1.1.4. Total IMC Thickness Measurement Analysis
4.1.1.5. Mechanical Ball Shear Test Analysis
4.1.2. Cu-Al Samples at Bonding Temperature of 280°C
4.1.2.1. FESEM Imaging Studies of Cu-Al Bonding Interface
4.1.2.2. EDX Compositional Studies of Cu-Al Bonding
Interface
4.1.2.3. STEM and EDX Analyses on Samples Before and
After 1000 hours of HTS Treatment
4.1.2.4. Total Cu-Al IMC Thickness Analysis
4.1.2.5. Mechanical Ball Shear Test Analysis
4.1.3. Cu-Al Samples Synthesized at Bonding Temperature of 400°C
94
94
95
95
v
80
82
83
87
89
90
93
101
106
113
114
115
115
122
127
137
138
139
4.2.
4.3.
4.4.
4.5.
4.6.
4.7.
4.8.
5
4.1.3.1. Morphological Studies of Cu-Al Bonding Interface by
FESEM
4.1.3.2. Compositional Studies of Cu-Al Bonding Interface by
EDX
4.1.3.3. STEM and EDX Analyses on Selected Cu-Al Samples
Synthesized at 400°C
4.1.3.4. Total IMC Thickness Measurement Analysis
4.1.3.5. Mechanical Ball Shear Test Analysis
Effect of Forming Gas Supply Towards The Micro-Structure of
Interfacial IMC
4.2.1. Cross-Sectional FESEM Imaging Analysis at Cu-Al Bonding
Interface
4.2.2. STEM and EDX Analyses at Cu-Al Bonding Interface
4.2.3. Mechanical Studies of Ball Shear Test
Mechanical Ball Shear Performance Analysis by Taguchi approach
STEM and EDX Analyses at Cu-Al Bonding Interface for Taguchi
Sample Matrix
4.4.1. Characterizations on Annealed Sample Synthesized at 100°C
Without Forming Gas Supply
4.4.2. Characterizations on Annealed Sample Synthesized at 280°C
With Forming Gas OFF
4.4.3. Characterizations on Annealed Sample Synthesized at 400°C
Without Forming Gas Supply
XRD Characterization With Two Different Sample Preparation
Techniques
4.5.1. By Mechanical Collection Technique
4.5.2. By Chemical Etching Technique
Theoretical Modelling
4.6.1. Prediction of Elastic Properties of Solid Solution
4.6.2. Prediction of Concentration of Solute and Partial Molar Volume
of Solid Solution
4.6.3. Prediction of Stresses
4.6.4. Stress Analysis to Evaluate the Effect of Bonding Temperatures
4.6.5. Stress Analysis To Assess the Effect of HTS Treatment
4.6.5.1. For Cu-Al Samples Synthesized at 100°C
4.6.5.2. For Cu-Al Samples Synthesized at 280°C
4.6.6. Stress Analysis to Evaluate the Effect of Forming Gas Supply
4.6.7. Stress Analysis Based on Taguchi Orthogonal Matrix
Applications of Taguchi Method on The Stresses Generated at Cu-Al
Bonding Interface
Summary
DISCUSSION
5.1. IMC Growth at The Bonding Interface and Its Correlation to The
vi
139
145
149
157
159
160
160
162
174
175
179
179
184
187
191
191
199
201
201
205
207
208
211
211
212
214
216
220
222
225
225
Mechanical Bond Strength
Relationship Between Ball Shear Strength and Bonding Parameters
Chemical Composition Distribution Across The Bonding Interface
Grain Boundary Diffusion at Low Bonding Temperature
The Presence of Oxide Layers at Cu-Al Bonding Interface
Formation of Gap Within The Oxide Layer at Cu-Al Bonding Interface
Possible Mechanism of The Formation of Uncommon Features at Cu-Al
Bonding Interface
5.8. Summary
5.2.
5.3.
5.4.
5.5.
5.6.
5.7.
6
CONCLUSION AND RECOMMENDATIONS
STUDIES
6.1. Conclusion
6.2. Contributions To Knowledge
6.3. Recommendations For Future Studies
FOR
229
230
232
233
237
240
245
FUTURE 247
247
248
249
251
264
REFERENCES
APPENDICES
vii
LIST OF TABLE
Table
Title
Page
2.1
Comparison between wire bonding technologies (Harman, 2010; Pan and
Fraud, 2004; Charles, 2007; Lum et al., 2006; Xu et al., 2008)
11
2.2
Material properties comparison between Cu and Au (Calister, 2004).
20
2.3
Micro-structural and physical properties of Cu-Al phases (Xu, 2010).
28
2.4
Wire bonding parameters applied in some of the reports focus on Thermosonic
Cu-Al system.
32
2.5
d hkl for various crystal systems and their relationship with Miller indices.
49
2.6
Orthogonal array (L9) (Puh et al., 2012)
56
2.7
S/N equations for different quality characteristic (Nalbant et al., 2007)
57
2.8
Standard two way AVONA arrangement (Sahai and Ageel, 2000).
59
2.9
Examples of critical F values based on α=0.10 (Walck, 2007).
61
3.1
Optimized Thermosonic wire bonding parameters used to synthesize Cu-Al
samples.
69
3.2
Full factorial sample matrix of Cu-Al samples.
71
3.3
Taguchi L9 Orthogonal Matrix used to synthesize Cu-Al samples.
72
3.4
Sample matrix of Cu-Al samples for direct comparisons.
73
3.5
Sample Matrix to evaluate the effect of HTS treatment on Cu-Al samples.
Forming gas supplies were turned ON.
74
viii
4.1
Summary of EDX measurements of sample synthesized at 100°C.
102
4.2
Summary of predicted Cu-Al phases from FESEM EDX for sample synthesized
at 100°C.
104
4.3
Summary of compositional analysis on as-synthesized sample produced with
100°C.
109
4.4
Summary of compositional analyses on sample synthesized at 100°C after 1000
hours of HTS.
112
4.5
Summary of EDX measurements of sample synthesized at 280°C.
123
4.6
Summary of predicted phases for sample synthesized at 280°C.
124
4.7
Summary of compositional analysis of Cu-Al bonding interface interpreted
from Figure 4.19.
130
4.8
Summary of compositional analysis of sample synthesized at 280°C after 1000
hours of HTS.
136
4.9
Summary of EDX measurements of sample synthesized at 400°C.
146
4.1
Summary of predicted phases for sample synthesized at 400°C.
147
4.11
Summary of compositional analysis of Cu-Al bonding interface of the sample
synthesized at 400° before HTS treatment.
152
4.12
Summary of compositional analysis of Cu-Al bonding interface of the sample
synthesized at 400° after 500 hours HTS treatment.
157
4.13
Summary of compositional analysis of the Cu-Al sample synthesized at 280°C
and without forming gas supply.
173
4.14
Results of ball shear measurements of the Cu-Al samples and their S/N ratio.
176
4.15
Response Table for S/N ratio.
176
4.16
Comparison between the predicted and measured S/N ratio and shear strength
of the Cu-Al samples.
177
4.17
ANOVA of ball shear strengths of the Cu-Al samples.
178
4.18
Summary of the compositional analysis on sample S4
183
4.19
Summary of the compositional studies on sample S5.
187
4.20
Summary of the compositional analysis of sample S9.
190
4.21
Summary of crystallographic detail of phase detected.
197
ix
4.22
Material related constants used in the analysis (Calister, 2004).
202
4.23
Results of the predicted stress and their S/N ratio
220
4.24
Response table for the S/N ratio of the stresses predicted.
221
5.1
Features and stresses of various locations at Cu-Al bonding interface of S1.
239
x
LIST OF FIGURE
Figure
Title
Page
2.1
Illustrations of (a) ceramic capillary, (b) cross-sectional view of the tapered tip
of (a). Modified from (CoorsTek, 2014)
12
2.2
Illustrations of (a) ball bond and (b) wedge bond. Modified from (Charles,
2009)
13
2.3
The illustrations for the process flow of Ultrasonic wedge bonding. Modified
from (Harman, 2010).
15
2.4
The illustration for the process flow of Thermosonic ball bonding. Modified
from (Harman, 2010).
18
2.5
Four tiered ball bond management in a micro-chip for graphic processing
applications (Singh et al., 2005).
20
2.6
The appearance of IMC layer in sample synthesized at 280°C after 500 hours
of HTS 175°C.
26
2.7
Binary Cu-Al thermal equilibrium phase diagram (ASM, 1992).
28
2.8
Comparison of FESEM micrograph of Cu FABs synthesized with (a) forming
gas ON, (b) forming gas OFF (Lee et al., 2010).
34
2.9
The design of a common copper kit. (Breach, 2010).
34
2.10
The construction of a conventional SEM equipment (Zhou et al., 2007).
37
2.11
SEM micrographs show the structures of (a) thermionic and (b) field effect
electron sources (Spectral solutions, 2014).
38
2.12
Illustration of signal generations resulted from interactions between incident
electron and the surface of a sample (Fultz and Howe, 2013).
40
2.13
Construction of TEM equipment (Fultz and Howe, 2013).
41
2.14
Schematic of an annular detector in a STEM. Modified from (Fultz and Howe,
2013).
43
2.15
The broadening of energetic electron beam in (a) bulk sample, and (b) in thin
lamella. Adapted from (Fultz and Howe, 2013).
45
xi
2.16
Illustration of EDX measurements taken at Cu, IMC and Al. Overlapping of
the interaction volume and Cu bulk results occurs when EDX measurement is
taken close to the Cu-IMC interface.
An indexed XRD diffractogram of Al thin film (Birkholz, 2006). Inset shows
the similar diffractogram with y-axiz in square-root of count rate.
45
2.18
Mechanical ball shear test configuration (Harman, 2010).
51
3.1
Flow chart of the processes of the experiment.
67
3.2
The illustration of the structure of the synthesized Thermosonic Cu-Al sample.
69
3.3
OM images of cross-sectioned samples with (a) a smooth polished bonding
interface, and (b) a damaged bonding interface.
77
3.4
The geometrical considerations of the multi-layer structure.
85
3.5
The illustration of the process flow of the stress calculation. (a) The
composition measurements, (b) the calculated x Cu profile from (a), (c) the
converted C profile, (d) Predicted elastic material properties, (e) predicted
stresses profile.
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 100°C before HTS treatment.
92
4.2
FESEM micrographs of cross-sectional Cu-Al bonding interface synthesized at
100°C after 200 hours of HTS treatment.
97
4.3
FESEM micrographs of cross-sectional Cu-Al bonding interface synthesized at
100°C after 500 hours of HTS treatment.
98
4.4
FESEM micrographs of cross-sectional Cu-Al bonding interface synthesized at
100°C after 700 hours of HTS treatment.
99
4.5
FESEM micrographs of cross-sectional Cu-Al bonding interface synthesized at
100°C after 1000 hours of HTS treatment.
100
4.6
(a) STEM overview of the lamella of sample synthesized at 100°C before HTS
treatment. The magnified micrographs at selected regions are shown in inset
(b) to (d).
(a) to (d) line scan EDX profiles (i) to (iv) as defined in Figure 4.6,
respectively.
108
4.8
NBD pattern taken from the IMC grain.
109
4.9
(a) STEM overview of the lamella of sample synthesized at 100°Cafter 1000
hours of HTS treatment. (b) to (d) show the magnified micrograph at selected
regions.
(a) to (c) were results of line scan EDX (i) to (iii) as defined in Figure 4.9,
respectively.
111
(a) The plot of total IMC thickness versus HTS durations for sample
synthesized at 100°C, (b) Results of linear regression analysis.
114
2.17
4.1
4.7
4.10
4.11
xii
48
96
108
111
4.12
Box plot of ball shear strength versus HTS durations for samples synthesized
at the bonding temperature of 100°C.
115
4.13
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C before HTS treatment.
117
4.14
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C after 200 hours of HTS treatment.
118
4.15
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C after 500 hours of HTS treatment.
119
4.16
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C after 700 hours of HTS treatment.
120
4.17
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 280°C after 1000 hours of HTS treatment.
121
4.18
(a) STEM overview of the lamella of the Cu-Al bonding interface synthesized
at 280°C before HTS treatment. (b) to (c) show the magnified micrographs at
selected regions.
(a) to (c) were line scan EDX Cu-Al profiles of measurements (i) to (iii) as
defined in Figure 4.18, respectively.
128
(a) STEM overview of the lamella of sample synthesized at 280°C after 1000
hours of HTS treatment. (b) to (d) show the magnified micrograph at selected
regions.
Magnified STEM micrograph revealed the existence of the particles and
stripes features in the IMC with darker. Synthesized conditions: bonding
temperature 280C and after 1000 hours of HTS treatment.
(a) to (f) were line scan EDX profiles (i) to (vi) as defined in Figure 4.20,
respectively.
131
4.23
NBD diffraction pattern that matched the diffration from zone axis [-1 -2 0] of
θ phase.
133
4.24
(a) The plot of total IMC thickness versus HTS durations for the Cu-Al
samples synthesized at 280°C, (b) Results of the corresponding linear
regression analysis.
Box plot of ball shear strength versus HTS durations for the Cu-Al samples
synthesized at the bonding temperature of 280°C.
137
4.26
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 400°C before HTS treatment.
140
4.27
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 400°C after HTS 200 hours.
141
4.28
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 400°C after HTS 500 hours.
142
4.29
FESEM micrographs of the cross-sectional Cu-Al bonding interface
synthesized at 400°C after HTS 700 hours.
143
4.30
FESEM micrographs of the cross-sectional bonding interface synthesized at
400°C after HTS 1000 hours.
144
4.19
4.20
4.21
4.22
4.25
xiii
128
132
133
138
4.31
(a) STEM overview of the lamella of sample synthesized at 400°C before HTS
treatment. (b) to (d) show the magnified micrograph at selected regions.
151
4.32
(a) to (c) show line scan EDX profiles of measurements (i) to (iii) as defined in
Figure 4.31, respectively.
151
4.33
(a) STEM overview of the lamella of sample synthesized at 400°C after 500
hours of HTS treatment. (b) to (d) show the magnified micrograph of the
selected regions around the bonding interface.
Results of EDX line scan measurements as defined by arrow (i) to (v) in
Figure 4.33.
153
(a) The plot of total IMC thickness versus HTS durations for Cu-Al sample
synthesized at 400°C, (b) the results of the corresponding linear regression
analysis.
Box plot of ball shear strength versus HTS durations for Cu-Al samples
synthesized at the bonding temperature of 400°C.
158
4.37
FESEM micrographs of the cross-sectioned of the as-synthesized Cu-Al
sample produced at 280°C and without the forming gas supply.
161
4.38
Bright field TEM micrograph of the Cu-Al sample synthesized at 280°C and
without forming gas supply. (a) to (d) are magnified micrographs that revealed
voids and IMC features. Inset (e) shows a magnified STEM micrograph that
better visualize the present of IMC.
OM images of (a) a symmetrical non-oxidized ball bond and (b) an
asymmetrical and off-centred oxidized ball bond.
163
Bright field TEM micrographs of Cu-Al bonding interface synthesized at
280°C and with forming gas supply ON. (a) to (d) are magnified micrographs
that revealed voids and IMC features at the bonding interface. (c) is a dark
field image.
IMC thickness comparison between sample with and without forming gas
supply.
167
4.42
Line scan EDX results across an IMC layer of (a) sample with forming gas
ON, (b) sample with forming gas OFF.
169
4.43
STEM micrographs of the as-synthesized sample bonded with bonding
temperature 280°C and forming gas OFF. (a) represents the overview
micrograph of the lamella. (b) to (d) are magnified micrographs at selected
region of the lamella.
Composition profiles across the Cu-Al bonding interface as defined by arrows
(i) to (vi) in Figure 4.43.
171
(a) and (b) are the FESEM micrographs for the Cu-Al bonding interface
synthesized with forming gas supply OFF and ON, respectively. Insets show
magnified micrograph in high stress regions for IMC thickness measurements.
The response graph of S/N ratio of ball shear strength with different factors
and levels.
174
STEM micrographs of the sample S4 synthesized with bonding temperature
100°C, forming gas OFF and after 500 hours of HTS treatment. (a) shows the
overview micrograph of the lamella. (b) to (d) are magnified micrographs at
181
4.34
4.35
4.36
4.39
4.40
4.41
4.44
4.45
4.46
4.47
xiv
153
159
165
168
171
177
some selected regions around the bonding interface.
4.48
(a) to (d) show the line scan EDX results measured along arrow (i) to (iv)
defined in Figure 4.47.
181
4.49
NBD diffraction patterns obtained from the IMC. The indexing of the
diffraction patterns matched (a) [1 -1 0] from θ, and (b) [-9 -7 6] from η 2
phases.
STEM micrographs of the sample S5 synthesized at 280°C with forming gas
OFF and after 500 hours of HTS treatment. (a) shows the overview micrograph
of the lamella. (b) and (c) are magnified micrographs at selected regions at the
bonding interface.
(a) to (c) show the line scan EDX results measured along arrow (i) to (iv) as
defined in Figure 4.50.
182
STEM images of sample S5 that was synthesized at 400°C without forming
gas supply and after HTS 1000hours. (a) was the overview micrograph, (b) to
(d) were the magnified micrographs at the selected RoIs.
Compositional studies on sample S9. (a) to (d) were the measurements’ results
taken along arrow (i) to (iv) as defined in Figure 4.52.
188
4.54
(a) The optical image of the wire bonded specimen in leadframe form. (b) The
schematic demonstrating the material configurations of a unit of specimen.
192
4.55
OM image of crushed specimen prepared by the manual collection of the wires
and micro-chips.
193
4.56
(a) FESEM of the crushed Cu-Al specimen, (b) corresponding EDX results
193
4.57
194
4.58
(a) Optical micrograph of a bundle of the specimen consisted of Cu wires and
Al bond pads. (b) magnified optical micrograph showing the existence of Al
under Cu ball bond.
EDX analysis on the bottom side of the ball bond. Al was detected.
4.59
XRD diffractograms for the specimens prepared with three different methods.
196
4.60
(a) The diffractogram from XRD measurement using Noiseless sample holder.
(b) A duplicate of diffractogram from Figure 4.59(b).
198
4.61
SEM EDX validation on the elimination of Cu-Al IMC after etching with
KOH 50%.
200
4.62
Diffractograms after KOH etching with different concentrations.
201
4.63
(a) The reference model of Cu-Al metal couple with a layer of solid solution
sandwiched between the pure metals. (b) The plot of Cu concentration versus
the thickness of the solid solution.
The comparison of reported and predicted values of (a) shear, (b) bulk and (c)
Young’s moduli.
202
4.50
4.51
4.52
4.53
4.64
xv
186
186
188
195
204
4.65
The plot of the predicted solute concentration, C versus fractional
concentration of Cu, x Cu .
206
4.66
The plot of the calculated partial molar volume of the solid solution, Ṽ versus
x Cu .
206
4.67
The calculated distribution of three components of the diffusion induced stress.
207
4.68
(a) STEM images on the IMC formation at the bonding interface. (b) The
corresponding x Cu profile. (c) The corresponding predicted stresses. (i) to (iii)
represent the results of the samples synthesized at 100, 280 and 400°C.
(a) STEM image of S7 synthesized at 100°C after 1000 hours of HTS
treatment. (b) x Cu profile obtained from the EDX measurement defined in (a).
(c) The predicted stress profile along the EDX line across the bonding
interface.
(a) STEM image of the bonding interface of S8 synthesized at 280°C after
1000 hours of HTS treatment. (b) x Cu profile calculated from the EDX
measurement defined in (a). (c) The predicted stress profile from x Cu profile.
(a) STEM image of the bonding interface of S1 synthesized at 280°C before
HTS treatment. (b) x Cu profile calculated from the line scan EDX measurement
defined in (a). (c) The predicted stress profile from x Cu profile.
(a) STEM micrograph of sample S4 synthesized at 100°C without forming gas
supply and after 500 hours of HTS treatment. (b) The calculated x Cu profile
from EDX measurement defined in (a). (c) The predicted stress distribution.
(a) STEM micrograph of sample S5 synthesized at 280°C without forming gas
supply and after 500 hours of HTS treatment. (b) The x Cu profile from the
EDX measurement defined in (a). (c) The calculated stress distribution.
STEM micrograph of sample S6 synthesized at 400°C with forming gas ON
and after 500 hours of HTS treatment. (b) the x Cu profile from the EDX
measurement specified in (a). (c) The predicted stress distribution.
STEM micrograph of sample S9 synthesized at 400°C with forming gas OFF
and after 1000 hours of HTS treatment. (b) the x Cu profile from the EDX
measurement specified in (a). (c) The predicted stress distribution.
IMC thickness versus HTS durations for the bonding temperatures of (a) 100,
(b) 280 and (c) 400°C.
209
5.2
Histogram of the ball shear strength distribution of samples prepared with
different temperatures and HTS durations.
228
5.3
The plot of ball shear strength versus total IMC thickness.
228
5.4
The chemical composition distribution at both central and peripheral RoIs.
232
5.5
Chemical compositional distribution at Cu-Al bonding interfaces of (a) S10,
(b) S8 and (c) S6.
234
5.6
(a) to (d) show the composition profiles of the bonding interface of samples
S1, S4, S5 and S9, respectively.
236
4.69
4.70
4.71
4.72
4.73
4.74
4.75
5.1
xvi
212
213
215
216
218
218
219
226
5.7
The compiled composition profiles and their respective stress profiles of
sample S1.
239
5.8
STEM micrograph of sample synthesized at 280°C with forming gas supply.
(a) and (b) were that before and after 1000 hours of HTS treatment,
respectively. Insets in (b) show the magnified micrographs that reveals the
presence of rod and stripe-like features.
STEM micrograph shows the appearance of stripes and rods-like feature
present at Cu-Al bonding interface.
241
Schematics of proposed mechanism of formations of uncommon features. (a)
the initial thin IMC formation, (b) with prolonged HTS, volume and lateral
diffusions of Cu atoms occurred simultaneously, (c) Upon consumption of the
bond pad, the diffused Cu atoms partially penetrated into Si together with
inversed diffusion to thicken a new IMC layer with higher Cu content, (d) the
formations of stripes, rods-like features and periodic Cu composition were
observed.
STEM micrograph of sample S6. The structure of the IMC formed was
apparently rods-like.
244
5.9
5.10
5.11
xvii
243
244
LIST OF APPENDICES
APPENDIX
A
B
C
TITLE
Derivations of ��0 , �� , ��0 And ��
Derivations of Conversion Equations
Derivation of The Number of Mole of Solute Atoms per
unit Volume, C
xviii
PAGE
264
269
272
LIST OF ABBREVIATIONS
Al
-
Aluminum
ASM
-
American Society for Metals
Au
-
Gold
BGA
-
Ball Grid Array
BST
Ball Shear Test
CTE
-
Coefficient of Thermal Expansion
Cu
-
Copper
EDX
-
Energy Dispersive X-Ray
EFO
-
Electro-Flame-Off
FAB
-
Free-Air-Ball
FESEM
-
Field Emission Scanning Electron Microscope
GB
-
Grain Boundary
HTS
-
High Temperature Storage
IMC
-
Intermetallic Compound
JEDEC
-
Joint Electronic Device Engineering Council
Ni
-
Nickel
SEM
Scanning Electron Microscope
Si
-
Silicon
STEM
-
Scanning Transmission Electron Microscope
W
-
Tungstun
XRD
-
X-Ray Diffraction
xix